[seweryn]

Hi,

I have 4 high resolution TSX images, and interested in do some sparse point processing with this too few images (I know that we need at least 15+ images for a robust PS processing), I started with small area processing to have a clue first, I used amp stab index+sp coherence to select points with a relatively high threshold 1.8 ( I am aware that with the few number of images the parameters have got higher values), when I try Multi-image Insar processing to estimate res.h and vel, the temporal coherence is significantly high, all above 0.95 and when I look at the time series of the selected points, they are all aligned in a straight line which sounds strange to me, it sounds like the observation points are drawing a straight line, which shouldn’t be true in reality, I can imagine with only 3 images (two ifgs) we can only fit a straight line and make a very high temp coherence, but when we increase the redundancy to 1 with having 4 images (3 ifgs) and two unknown why the observations are again aligned in a straight line? I also can see a message say the that the master error skipped to correct because of too few images. Do you have any reason for that to let me know, or any suggestion for a few images to process in SARPROZ?

Thanks

Attachments:

You must be logged in to view attached files.

[periz]

you are not fitting a single line. you are fitting a planar surface (height and velocity).
with 4 images you can look at 3 interferograms. or you can calculate the average if in your particular case it makes sense. or you can neglect height and atmosphere if you are in the proper situation and assume that the phase content is just movement.
in general, just calculate interferograms and check them.
best

[Author]

Posts